Postoperative adverse neurological symptoms are frequent occurrences in the older patients after major surgery. These involve delirium, subsyndromal delirium and opioid toxicities. An acute change in mental status, along with inattention are often hallmark of a larger geriatric syndrome commonly called postoperative delirium, and a related entity, subsyndromal delirium can be considered as subclinical delirium. Patients who develop postoperative delirium typically stay in the hospital longer and have an increased risk of post- discharge decline in functional and cognitive status, and even increased long-term mortality. Patients with opioids toxicities such as over-sedation may also have associated respiratory depression, and if not clinically recognized, may progress to full respiratory arrest if the depression is profound and prolonged. The symptoms of postoperative delirium, subsyndromal delirium, and opioid toxicities are commonly thought to be related to the administration of intravenous opioids, conventional therapy prescribed for the alleviation of postoperative pain. For patients who have these adverse neurological symptoms, the typical strategy is to stop the administration of opioids in the hope of reversing these changes. However, with the cessation of opioids administration, patients may rebound with substantial postoperative pain which may actually increase their symptoms of confusion since pain has been reported to be associated with postoperative delirium. In current clinical practice, the titration of intravenous opioids in the treatment postoperative pain is largely empirical. Whether biomarkers can explain, or predict the symptoms of postoperative cognitive changes has not been previously evaluated in a systematic fashion. The objective of this study is to leverage several existing large patient cohort databases to determine the relationship of genetic variants and the phenotypes of postoperative delirium, subsyndromal delirium and opioid toxicities. The combined databases to be used in this investigation consist of over two thousand older patients who have undergone major non-cardiac surgery at three separate university medical centers. All patients were assessed for the presence of in-hospital postoperative delirium, subsyndromal delirium, and opioid toxicities using validated and structured protocols. Because three heterogeneous patient populations will be included in the discovery cohort plus a 4th prospectively recruited cohort as the validation cohort, our study has the unique opportunity to determine both internal and external validity of our findings. Our central hypothesis is that individuals who have a certain genetic profile have a higher incidence of postoperative delirium, subsyndromal delirium, and opioid toxicities. Our results will have significant impact on precision medicine and our strategy of assessment of biomarkers to facilitate personalized postoperative pain and symptom management may result in a decrease of postoperative delirium, subsyndromal delirium, and opioid toxicities in the at risk older population.